delivery of high-quality biomarker assaysdownloads.hindawi.com/journals/dm/2002/212987.pdf ·...

Disease Markers 18 (2002) 47–56 47IOS Press

Delivery of high-quality biomarker assays

Brian N. Swanson∗Clinical Discovery, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08690, USA

Abstract. Biomarker measurements now support key decisions throughout the drug development process, from lead optimizationto regulatory approvals. They are essential for documenting exposure-response relationships, specificity and potency toward themolecular target, untoward effects, and therapeutic applications. In a broader sense, biomarkers constitute the basis of clinicalpathology and laboratory medicine. The utility of biomarkers is limited by their specificity and sensitivity toward the drugor disease process and by their overall variability. Understanding and controlling sources of variability is not only imperativefor delivering high-quality assay results, but ultimately for controlling the size and expense of research studies. Variability inbiomarker measurements is affected by: biological and environmental factors (e.g., gender, age, posture, diet and biorhythms),sample collection factors (e.g., preservatives, transport and storage conditions, and collection technique), and analytical factors(e.g., purity of reference material, pipetting precision, and antibody specificity). The quality standards for biomarker assaysused in support of nonclinical safety studies fall under GLP (FDA) regulations, whereas, those assays used to support humandiagnostics and healthcare are established by CLIA (CMS) regulations and accrediting organizations such as the College ofAmerican Pathologists. While most research applications of biomarkers are not regulated, biomarker laboratories in all settingsare adopting similar laboratory practices in order to deliver high-quality data. Because of the escalation in demand for biomarkermeasurements, the highly-parallel (multi-plexed) assay platforms that have fueled the rise of genomics will likely evolve into theanalytical engines that drive the biomarker laboratories of tomorrow.

Keywords: Assay, bioanalytical, biomarker, CLIA, drug development, GLP, pharmaceutical, quality, regulations, review, valida-tion

1. Introduction

Heightened focus on the use of bioimaging andex vi-vo laboratory measurements to guide the developmentof new therapies is transforming processes and expec-tations within the pharmaceutical research and regu-latory communities [2]. Drug developers are findingnew ways of supporting clear decision-making, limit-ing downstream risk of project failure, and document-ing therapeutic claims. Earlier assessment of the ther-apeutic benefits and liabilities of new chemical enti-ties promises substantial improvements in the overallcost and time for drug development. Regulators areenthused by the prospect of a richer array of informa-tion supporting the safety and efficacy of new drug en-

∗Address for correspondence: Dr. Brian N. Swanson, Director,Clinical Laboratory, Bristol-Myers Squibb Company, 3 HamiltonHealth Place, Hamilton, NJ 08690, USA. Tel.: +1 609 689 4567;Fax: +1 609 689 4250; E-mail: [email protected].

tities. And there is renewed enthusiasm and expecta-tion for the use of surrogate clinical endpoints whichcan assess the effects of therapeutic intervention morequickly than conventional observations of morbidityand mortality.

The impetus for these sweeping changes are two-fold. Although biological markers (biomarkers)and associated bioanalytical tools have always beenthe bedrock of pharmacology and toxicology, olderparadigms for drug development cannot be sustaineddue to ballooning costs, especially during the clinicalphases. The biomarkers of tomorrow must becomemore timely, reliable and definitive to propel decision-making. Drug candidates that ultimately fail must doso far earlier in the process. Secondly, revolutionarychanges in both our understanding of human genomicsand bioanalytical technologies have enabled the iden-tification and quantitation of a wide array of biomark-ers. This knowledge and technology will allow com-prehensive profiling of individual research subjects and

ISSN 0278-0240/02/$8.00 2002 – IOS Press. All rights reserved

48 B.N. Swanson / Delivery of high-quality biomarker assays

Personalized Therapy in Practice

Rational explanation of: • variability in drug response • variability in PK/ADME • adverse events • alleles vs drug response • biomarker vs disease status

Prognostic & diagnostic tests for pharmacotherapy (theranostics)

Profiling of Research Subjects

• Biomarker Assays • Clinical Lab Assays • Genetic Profiling • Expression profiling • Proteomic profiling • Drug PK

Comprehensive analysis of genotype/pathophysiol. status

Fig. 1. Projected impact of comprehensive clinical laboratory testing on pharmacotherapy.

inevitably drive medical practice toward personalizedtherapy that is less empirical and driven by laboratorydata (Fig. 1).

The challenge that drug researchers and bioanalyti-cal chemists face today are practical ones having to dowith the selection of appropriate biomarkers for criticaldecisions, the delivery of high-quality biomarker meth-ods, and the evaluation of the data. All of these pro-cesses are contingent upon clearly understanding andcontrolling the factors that influence the reliability andvariability of the data [12]. Indeed, the size of studygroups (e.g., in toxicology studies and clinical trials)and statistical plans are a direct function of variabilityin the measurements. Analytical sources of variationcan be reduced by the use of rugged, validated assaymethodologies. Biological sources of variation can becontrolled in part by experimental design (e.g., by spec-ifying inclusion and exclusion criteria, restricting foodconsumption, defining the time of day for specimencollections).

This paper summarizes the various bioanalytical andbiological factors that influence the quality and vari-ability of ex vivo biomarker measurements. It beginswith a consideration of the key guidelines and regula-tions that have molded approaches to assay validationand then presents practical recommendations that willassure consistent delivery of high-quality biomarkerdata.

2. Regulatory considerations

Bioanalytical laboratories that support pharmaceuti-cal research and development are often categorized ac-cording to their level of compliance with the federalreg-ulations which govern animal toxicology studies (GoodLaboratory Practices; GLP) and human diagnostic test-ing (Clinical Laboratory Improvement Amendments;

CLIA). GLP regulations are enforced by the US Foodand Drug Administration (FDA) and were developed toassure the quality of data generated for toxicology andsafety pharmacology studies in animals [3]. They donot apply to most of the exploratory work done in ani-mal pharmacology laboratories. CLIA regulations areadministered by the Centers for Medicare & MedicaidServices (CMS), formerly known as the Health CareFinance Administration [4]. The purpose of CLIA reg-ulations is to ensure the quality of the assay work per-formed in “any facility which performs laboratory test-ing on specimens derived from humans for the purposeof providing information for the diagnosis, prevention,treatment of disease, or impairment of, or assessment ofhealth”. Although CMS administers CLIA regulations,the FDA has an important role in regulating the testingin clinical laboratories by reviewing and approving thecommercial reagents and kit methodologies used fordiagnostic purposes. The CLIA rules do not apply to“research laboratories that test human specimens but donot report patient specific results for the diagnosis, pre-vention or treatment of any disease or impairment of,or the assessment of the health of individual patients”.FDA regulations for Good Clinical Practices (GCP),which govern the testing of new medical devices in hu-mans, do not specifically apply to clinical laboratories,but the data generated by clinical laboratories on speci-mens derived from clinical drug trials must be present-ed in final study reports and adverse event documentssubmitted to the FDA.

The recent escalation in biomarker measurementsthroughout drug development has brought about a con-vergence of laboratories from GLP-regulated, CLIA-regulated, and basic research settings. In this new en-vironment, bioanalysts are prone to confuse the appli-cability of GLP and CLIA regulations (and their regu-latory counterparts outside the USA). Table 1 summa-rizes some key similarities and differences. Both GLP

B.N. Swanson / Delivery of high-quality biomarker assays 49

Table 1Comparison of GLP and CLIA regulatory requirements

Requirement GLP CLIA

Overall scope Nonclinical drug safety studies (animalsonly)

Human diagnostics and prognostics

Applicable to basic research? No NoFDA-regulated Yes Yes, commercial reagentsCMS-regulated No YesLaboratory accreditation No YesLaboratory license No YesPersonnel qualifications No specific requirements Detailed requirementsWritten SOPs Yes YesEquipment maintenance and calibration SOPs andrecords

Yes Yes

Reagent identity labeling and expiration dates Yes YesApproved written study plan/protocol Yes No, but required by FDA GCP for clini-

cal trialsMandatory external proficiency testing No YesInternal compliance audits vs study protocol/plans Yes NoInternal compliance audits vs SOPs and regulations Yes, per each study Yes, annually for all operationsExternal compliance inspections Yes, by FDA Yes, by CMS and by state authoritiesFinal report for each study Yes Discretionary for lab, but mandatory for

GCP studiesAudit trail for data Yes YesRetention of records Shorter of: 5 years after NDA submis-

sion, 2 years after discontinuation, or 2years after FDA approval to market

5 years for routine diagnostics; FDA lim-its apply for GCP studies

and CLIA regulations establish quality standards forfacilities, personnel and procedures. However, GLPregulations do not apply to studies involving humansubjects, so, by definition, human specimens cannot beprocessed “under GLP”. At the same time, it is recog-nized that “GLP-like” practices will add an element ofquality to work performed on clinical specimens. Sim-ilarly, CLIA regulations do not apply to animal safetystudies and may not even apply to clinical biomark-ers when the tests are purely of a research nature (i.e.either of unproven value in guiding patient care deci-sions or not reported to participating human subjectsand their attending physicians). Even so, running theseresearch assays using “CLIA-like” practices will addto the quality and credibility of the data.

This raises two questions. First, to what extentshould biomarkers that are not being used for nonclin-ical safety studies or patient care be measured withinstrict GLP-like and/or CLIA-like environments? Theshort answer is that the quality standards embodiedby GLP-like and/or CLIA-like practices are impor-tant for all biomarkers used to support critical deci-sions. Implementation of best laboratory practices en-sures that biomarker laboratories will deliver an arrayof high-quality assay results, irrespective of the reg-ulatory environments under which the specimens arecollected. This basically means operating under stan-dardized written guidelines for assay validation, sam-

ple processing, quality control, and data management.Secondly, how onerous is it for biomarker laboratoriesto establish and maintain GLP-like and/or CLIA-likeconditions? While research staff will often eschewmore rigid processes due to the perceived investment ofadditional time and expense, the reality is that qualitysystems are relatively straightforward to maintain andwill reimburse the initial investment in Standard Oper-ating Procedures (SOPs) and other documentation byavoiding errors in sample collection and processing.

3. Quality standards for analytical validation andlaboratory practices

The original GLP regulations were not specific asto requirements for assay validation and quality con-trol. Consequently, bioanalytical laboratories support-ing preclinical studies developed a broad range of ap-proaches to assay validation. In 1990, representativesof various bioanalytical laboratories associated with thepharmaceutical industry met to develop a set of uni-form guidelines for drug assay validation [41]. Theseguidelines were widely adopted for GLP studies [8,18,19]. Because of their focus on chromatography meth-ods and the subsequent rise of biotechnology prod-ucts dependent on immunoassay technology, addition-al guidelines have been more recently developed to


encompass immunoassays [11,42]. Regulatory guide-lines for validation of drug assays have now been issuedby the FDA and The International Conference on Har-monisation of Technical Requirements for Registrationof Pharmaceuticals for Human Use (ICH) [16,23,24].While none of these guidelines specifically apply tobiomarker assays, their principles of quality have clearapplication. One obvious shortcoming is that they donot consider the broad range of biological factors thatimpact biomarker assay validation.

Quality standards for clinical laboratories in the USAhave their roots in the Clinical Laboratory ImprovementAmendments (CLIA) and in accrediting organizationssuch as the College of American Pathologists (CAP)and the Joint Commission on Accreditation of Health-care Organizations (JCAHO) [17]. Accrediting orga-nizations adopt guidelines and polices that go beyondthe generalities of regulatory requirements and pro-mote specific, best laboratory practices across all pro-cesses and require well defined quality assurance pro-grams. Thus, accreditation by CAP or JCAHO assuresfull compliance with CLIA regulations plus achieve-ment of an overall high standard of laboratory prac-tice. These regulations and guidelines require all assaymethods to be periodically validated for accuracy andreproducibility, to be checked daily for quality controlperformance, to have biological reference ranges de-fined, and, for regulated analytes, to be monitored viaexternal proficiency testing. The National Committeefor Clinical Laboratory Standards (NCCLS) promotesbest practices by developing standard operating proce-dures in collaboration with teams of experts and issu-ing SOPs for a wide range of clinical laboratory proce-dures. Among the NCCLS publications are guidelinesfor assessing assay performance and analytical valid-ity. Similar bioanalytical performance standards havebeen developed in Europe and internationally [7,14,22,26,30,40,44]. In general, these various guidelines areapplicable across all assay methodologies; however,some technologies, such as flow cytometry and quan-titative NMR, pose special challenges for validationand these have been addressed in additional publica-tions [35,36]. Probably no single individual has donemore to advance laboratory quality control systems andtechniques than James O. Westgard [49]. A wealth ofpractical information on assay validation experimentsand approaches to quality control can be found at hiswebsite Westgard.com.

4. Sources of variability in biomarkermeasurements

By their nature, biomarkers must be dynamic andresponsive to the disease process or pharmacologicalintervention of interest. Imposed upon these desir-able signals is the background noise of biological andmethodological sources of variation [12,37]. Thesebackground fluctuations affect both the sensitivity andthe specificity of biomarker measurements and ulti-mately limit their utility. The size of treatment groupsneeded for drug studies is directly related to the variabil-ity in the measurements used for endpoints and deci-sions. Practical implementation of biomarkers in drugdevelopment therefore requires both an understandingand control of the various sources of variability in assayperformance [27].

Preanalytical sources of variability in biomarkermeasurements are noted in Table 2 [1,6,10,13,25,29,31–33,46,47]. Narayanan has recently reviewed manyof these variables [34]. The impact of these factors isoften not well understood at the time new biomarkersare implemented in drug studies. This is especiallytrue for those biomarkers specifically linked to novelmolecular targets. To deliver high-quality biomarkerresults, drug researchers must either use study inclu-sion and exclusion criteria that control these factors orelse undertake pilot studies that seek information onthese factors. For example, as part of the biomarkerassay validation process, reference ranges can be es-tablished using specimens from healthy individuals aswell as from subjects with the targeted disease. Also,small numbers of subjects can be evaluated as inpa-tients for a limited period of time to define variabilitydue to diurnal rhythms, posture, exercise, and meals.

Timed urine collections are particularly problematicin that subjects can fail to void completely either priorto, or at the end of, the collection interval of interest orsimply discard urine that is voided in the middle of thecollection interval. To minimize the impact of thesecollection errors, normalization of quantitative urinarybiomarkers using urinary creatinine is highly recom-mended. The rate of creatinine production is relativelyconstant for an individual, although it can vary some-what immediately following ingestion of well-cookedmeat [45].

Analytical sources of variability in biomarker mea-surements are summarized in Table 3. Standardiza-tion of biomarker assays in a research environment isparticularly challenging, given that standard referencematerials are often not readily available. Purification


Table 2Preanalytical sources of variability in biomarker measurements

Biological Sample collection

Genetics/gender MislabelingAge Duration of tourniquet applicationPosture Strength of collection vacuumExercise Size of needle gaugeMeals/fasting status Dead volume in catheters/collection tubesDiet AnticoagulantsDiurnal biorhythm Local effects of indwelling catheterSeasonal biorhythm Time and temperature prior to centrifugationConcurrent diseases Centrifugation speed, duration, temperatureConcurrent medications Evaporation/biomarker volatilityOverall health/preexisting disease Preservatives/biomarker instabilityGI motility Storage temperatureAnesthesia/surgical intervention Transport temperatureStress Completeness of urine collectionPregnancy HemolysisMenstrual cycle Exposure to lightDehydration

Table 3Analytical sources of variability in biomarker mea-surements

Purity of reference standardsLot-to-lot variation in reagentsPipetting imprecisionAntibody cross-reactivityLoss during extractionCross-contamination of pipetsDilution errorsMislabeling of processing tubesPre-assay incubation time & temperaturePre-assay amplificationsChemical interference by endogenous compoundsChemical interference by drugsAnalyte or reagent instability in lightTime between intermediate stepsFluctuations in instrument performanceCorrection for baseline/background levelsPost-run calculation errors

of endogenous biochemicals and enzymes may not befeasible without loss of essential biological or chemicalproperties. Also, biomarker assays can be a functionof a complex series ofin vivo andex vivo proceduresand measurements. For example, cell incubation steps(such as lipopolysaccharide stimulation) and amplifi-cation steps (such as PCR) can be integral parts of thebiomarker assay method. To correct for variability in-troduced by pre-assay manipulations of cells or bio-logical matrix, it may be necessary to include internalchemical controls or normal control samples that areused to normalize the final biomarker results.

Commercial reference materials for large peptidesand proteins can be particularly problematic becausethe assigned potencies can be based on biological testsystems (rather than chemical properties), whereas thecorresponding biomarker assays can report results in

terms of absolute concentration (e.g., ng/mL of plas-ma). This is frequently true for cytokines. The bioana-lyst might assume that a weighed amount of referencematerial has the same biological potency (and purity)across lots, but this cannot be assured. While absoluteaccuracy is not necessarily essential in an unregulatedresearch setting, every effort should be made to employreference standards of documented and reproduciblepurity so that assay results from different experimentsand from different analytical laboratories can be reason-ably compared. This highlights the great importanceof utilizing long-term quality control samples that areindependent of the calibrators and that verify consis-tent assay performance across transitions in standardsand reagents. When using commercial quality controlmaterial, each laboratory should establish a mean valuebased on actual assay performance in the local labo-ratory, rather than accepting the manufacturer’s statedvalue. Indeed, most suppliers of quality control materi-al no longer state an absolute value, but rather, a targetrange.

One of the most intuitively obvious, yet frequent-ly overlooked, sources of analytical error is the inter-ference of the test drug with the biomarker reagentsand detection systems. It should be presumed that allnew drug entities can potentially interfere with the per-formance of biomarker assays, for example, by alter-ing absorbance and fluorescence endpoints, co-elutingin chromatographic systems, or cross-reacting withreagent antibodies [50]. Therefore, all bioanalyticallaboratories supporting a drug research project shouldverify that the drug does not interfere with the applica-ble biomarker and diagnostic tests. This is most direct-ly done by assessing the performance of the biomarker


assay before and after spiking the drug (and, if avail-able, known drug metabolites) into the biological ma-trix of interest. Various endogenous substances canalso potentially interfere with biomarker assays [38].

5. Minimum requirements for biomarker assayvalidation

Given the fact that most biomarkers in a research set-ting are not subject to GLP and CLIA regulations, whatis the extent of validation needed for measurements thatdrive drug development decisions? While no consen-sus paper has been issued, it is generally recognizedthat sound assay validation should include documenta-tion of the parameters listed in Table 4. These conceptscan be illustrated by validation experiments that wereundertaken for a kit enzyme immunoassay for IL-2 sol-uble receptor alpha (Quantikine� ELISA, R&D Sys-tems, Minneapolis, MN). The validation data providedin commercial kit documentation should not be takenat face value because assay performance can be de-pendent upon personnel, equipment, working environ-ment, reagent water and other local factors (Table 3).Also, commercial kits that are marketed for researchpurposes only are not subjected to the rigorous FDAreview that is mandatory for diagnostic reagents. GoodManufacturing Practices are not necessarily utilized bycommercial vendors, resulting in marked variability inlot-to-lot kit performance.

Fundamental to all quantitative assays is a document-ed concentration-response relationship, preferably us-ing a well-characterized reference material. This isusually shown as a fitted curve based on responses ob-served for a biological matrix or buffer spiked witha pure reference standard (Fig. 2). During the assayprocess, responses observed for biological samples arereferred to the standard curve to derive the concentra-tions. The goodness of fit for the curve-fitting function,and hence the ability of the calibration curve to reportaccurate results throughout its concentration range, ischecked by “reading back” the responses of standardsagainst the curve to arrive at theoretical measured con-centrations for the standards. In general, the read-backvalues should be within a few percentage points of truevalue throughout the functional range of the assay (Ta-ble 5).

If a calibration curve is used in place of a standardcurve in the biological matrix, then the ability to ac-curately recover the analyte from the matrix must beestablished. This is done by spiking known amounts of

0.010

0.100

1.000

10.000

10 100 1000 10000

IL-2sR alpha (pg/mL)

O.D

. (m

ean

+/-S

D)

Fig. 2. Calibration curve for IL-2 soluble receptor alpha in bufferusing an ELISA kit from R&D Systems. Dilutions of a concentratedkit standard were prepared in a diluent supplied in the kit. Opticaldensities were determined in the ELISA plates for 5 assay runs and themean values plotted versus the concentration using a four-parameterlogistic fit.

y = 0.99x + 594

R2 = 0.999

Mean Recovery = 102%

0

2000

4000

6000

8000

10000

12000

14000

0 5000 10000 15000

Added IL-2sR alpha (pg/mL)

Fig. 3. Analytical recovery of recombinant IL-2 soluble receptoralpha when added to human plasma and assayed with an ELISA kitfrom R&D Systems. Percent recovery was calculated after subtract-ing the amount of endogenous material measured in the unspikedsample.

pure reference material into the matrix (covering a wideconcentration range), assaying the spiked samples, andthen calculating the recovery, taking into account theoriginal amount of endogenous analyte present in thematrix (Fig. 3). Since immunoassays in particular havea discrete, limited functional range, one should verifythat samples with concentrations higher than the high-est functional standard can be diluted into the function-al range and assayed with good accuracy (after correc-tion for the dilution factor). The analysts should know


Table 4Minimum requirements for biomarker assay validation

Reference standard of known purity or specific activityDefined concentration-response relationship (standard curve or calibration curve)Accurate curve-fitting function (accurate °read-back– of standards)Accurate recovery from biological matrixAccurate recovery for samples requiring pre-assay dilutionHighest concentration not requiring dilutionFunctional limits of the calibration or standard curveWithin-assay precision (CV) for at least 3 samples, 6 measurements eachBetween-assay precision (CV) for at least 3 samples, 6 runs eachStability during maximal time needed for each step (sample collection, transport, storage and intermediate steps)Proof of specificityLack of chemical/immunochemical interference by test drugsComparison with previous method (when possible)Interlaboratory comparison with blinded samples (when possible)Reference range for pertinent populationQuality controls independent of calibratorsExplicit rules for acceptance/rejection of assay runs

Table 5Read-back values in assay buffer for IL-2 soluble receptor alpha calibrators versus nominal(true) concentrations using an ELISA kit from R&D Systems. After a typical calibration curvewas generated using a four-parameter logistic fit, read-back concentrations for the calibratorswere derived from the fitted curve. The mean results for 6 different assay runs are shown

Nominal value (pg/mL) Mean read-back value (pg/mL) SD CV % of Nominal

78 65 13 20% 84%156 161 13 8% 103%313 325 23 7% 104%625 629 44 7% 101%

1250 1231 101 8% 98%2500 2522 234 9% 101%5000 5042 783 16% 101%

the upper limit of the functional range; all sampleswith initial values above this limit must be diluted andre-assayed.

Standards that are based on the biomatrix and aretaken through all assay steps, automatically correct foranalyte losses (e.g., during extraction or evaporationsteps). In contrast, if calibrators in buffers are used,then a factor accounting for percent loss may be neededto correct the results for biological samples that passthrough all preparatory steps. Failure to correct theresults for these losses is not uncommon in laborato-ries measuring biomarkers; consequently, marked dif-ferences in absolute concentrations can occur betweenlaboratories. Variation in the purity or specific activityof the reference standard can also account for inter-laboratory differences. To assure that assay results canbe reasonably compared between laboratories and/orbetween analytical runs, each laboratory should indi-cate whether or not assay results are corrected for re-covery and should declare the purity or specific activityof the reference standard in use. As an overall check onaccuracy, it is recommended that samples be split and

assayed simultaneously at another laboratory alreadyperforming the assay.

Overall precision should be checked both within andacross multiple runs (Table 6). To assure that an as-say continues to perform consistently across multipleruns, quality control samples based on the biologicalmatrix should be prepared, preferably at three differentconcentrations, and assayed with each run. Clear rulesfor accepting or rejecting analytical runs based on QCoutcomes must be stated (e.g., 2 of 3 within 15% ofnominal value, or 2 of 3 within 2 SD of assigned meanwith SD based on interassay CV).

Potential interference from endogenous compoundsand the test drugs should be considered and assessedas needed. The specificity of reagent antibodies shouldalways be established relative to compounds of simi-lar structure. For example, R&D Systems establishedthat the antibodies used in its kit for assay of IL-2 sol-uble receptor alpha did not cross-react with 71 othercytokines and chemokines.

Stability of the analyte should be verified for everystep of the assay process where timing is not absolute-ly controlled. This can apply to specimen collection,


Table 6Assay precision for IL-2 soluble receptor alpha in plasma using an ELISA kit from R&D Systems. Human plasmawas supplemented with recombinant IL-2 soluble receptor alpha and was assayed repetitively within the same assayrun (N = 6 measurements) and between multiple runs (N = 18 runs)

Sample Within-assay mean (pg/mL) Within-assay CV Between-assay mean (pg/mL) Between-assay CV

1 1,165 2.5% 1,093 4.7%2 3,864 6.8% 3,419 8.8%3 10,401 6.7% 9,193 5.2%

0

2000

4000

6000

8000

10000

0 10 20 30

Days of storage at -70 C

IL2-

sRa

(pg

/mL

)

Fig. 4. Stability of IL-2 soluble receptor alpha in plasma stored at−70◦C. Three different plasma samples were divided into multiplealiquots which were stored in the same freezer and later thawed andassayed at selected times.

transport, storage, and intermediate steps in samplepreparation, such as extractions and freeze-thaw cycles.Ideally, stability experiments should cover the longesttime envisioned for these steps, but this is not alwayspractical (Fig. 4).

6. Conclusion

Completion of the human genome sequence has ini-tiated a rapid proliferation in the number of molecu-lar targets for novel drug therapies and has fueled thesearch for downstream markers that reflect disease pro-cesses and therapeutic interventions [9,15]. Strategicuse of biomarkers can enable the evaluation of greaternumbers of drug candidates, while potentially decreas-ing the overall time and expense by eliminating infe-rior compounds earlier in the selection process. Asthe relationships between the genotypes, various mark-ers and therapeutic outcomes are discerned, an arrayof clinically-validated diagnostic, prognostic and ther-anostic assays will emerge [20,21]. Thus, it is likelythat comprehensive profiling of both the drugs in de-velopment and the patients undergoing treatment willbe essential for successful outcomes [28,39].

Juxtaposed in the midst of these revolutionarychanges in drug development and laboratory medicineare bioanalytical laboratories that must deliver high-

quality data supportive of accurate and timely decision-making. In order to launch rugged and reproduciblebiomarker methods, the assay validation processesshould account for all significant sources of variation.While analytical factors related to assay methodologyand performance can be readily understood and con-trolled in the laboratory, the biological and pathophys-iological factors may require exploratory studies in-volving collection of specimens from donors with thetarget disease under controlled conditions (e.g., controlof diet, posture, time of collection). For biomarkersused as therapeutic endpoints, this kind of informationis especially important for planning clinical drug tri-als (e.g., definition of the inclusion and exclusion cri-teria, specimen collection schedules, and numbers ofsubjects required for statistical assessment of efficacy).

Collectively, these impending changes in researchand patient care paradigms will compel research, toxi-cology and clinical laboratories to deliver high-qualitybioanalytical data in amounts far beyond what is cur-rently possible. The autoanalyzers of today are stillvariants on serial processing, albeit progressing inspeed and diversity of test menus. Batch, single-assay processing (e.g., EIA) can be faster but lacksthe random-access capabilities needed for rapid cus-tomization of test menus and cannot readily supportthe kind of multi-analyte profiles that will be required.To cope with these demands, bioanalysts must nowturn to the newer technologies that offer highly-parallel(“multiplexing”) capability, several of which have fu-eled genomics research [5,43,48]. Transformation ofmicrochip platforms into analytical systems of superi-or accuracy and precision will be challenging but like-ly. Finally, special bioinformatics tools will be need-ed to translate biomarker profiles into practical indicesthat can be readily interpreted and used to acceleratedecision-making.

Acknowledgements

The assay validation data for IL2 soluble receptor al-pha were generated by Alex Bragat and Ole Vesterqvist


of the Bristol-Myers Squibb Company. Several of theconcepts underlying biomarker assay validation weredeveloped in collaboration with Dr. Vesterqvist.

References

[1] N. Arvidson, B. Gudbjornsson, L. Elfman, A.-C. Ryden,T.H. Totterman and R. Hallgren, Circadian rhythm of seruminterleukin-6 in rheumatoid arthritis,Ann Rheum Dis 53(1994), 521–524.

[2] Biomarkers and Surrogate Endpoints: Advancing Clinical Re-search and Applications, A Multidisciplinary, InternationalConference Sponsored by NIH and FDA, Natcher ConferenceCenter, National Institutes of Health, Bethesda, MD, April15–16, 1999.

[3] Code of Federal Regulations, Title 21, Volume 1 [Revised asof April 1, 2001], US Government Printing Office via GPOAccess, 21CFR58, Good Laboratory Practice for NonclinicalLaboratory Studies.

[4] Code of Federal Regulations, Title 42, Volume 3 [Revisedas of October 1, 2001], US Government Printing Office viaGPO Access, 42CFR493, Clinical Laboratory ImprovementAmendments.

[5] L. Bubendorf, High-throughput microarray technologies:from genomics to clinics,Eur Urol 40 (2001), 231–238.

[6] L.M. Burrell, J. Palmer, J.A. Charlton, T. Thomas and P.H.Baylis, A new radioimmunoassay for human alpha atrial natri-uretic peptide and its physiological validation,J Immunoassay11 (1990), 159–175.

[7] D. Burnett and C. Blair, Standards for the medical laboratory –harmonization and subsidiarity,Clin Chim Acta 309 (2001),137–145.

[8] R. Causon, Validation of chromatographic methods in biomed-ical analysis. Viewpoint and discussion,J Chromatogr BBiomed Sci Appl 689 (1997), 175–180.

[9] A. Cruvchik, G. Subramanian, S. Bruder and J.C. Venter,Sequence analysis of the human genome: implications for theunderstanding of nervous system function and disease,ArchNeurol 58 (2001), 1772–1778.

[10] I.B. Dugue, E. Leppanen and R. Grasbeck, Preanalytical fac-tors and the measurement of cytokines in human subjects,IntJ Clin Lab Res 26 (1996), 99–105.

[11] J.W. Findlay, W.C. Smith, J.W. Lee, G.D. Nordblom, I. Das,B.S. DeSilva, M.N. Khan and R.R. Bowsher, Validation ofimmunoassays for bioanalysis: a pharmaceutical industry per-spective,J Pharm Biomed Anal 21 (2000), 1249–1273.

[12] C.G. Fraser,Biological Variation: From Principles to Prac-tice, AACC Press, Washington, DC, 2001, pp. 151.

[13] B.J. Gertz, J.D. Clemens, S.D. Holland and S. Greenspan,Application of a new serum assay for type I collagen cross-linked N-telopeptides: assessment of diurnal changes in boneturnover with and without alendronate treatment,Calc Tiss Int63 (1998), 102–106.

[14] D.J. Goldie, Accreditation of external quality assessmentschemes in the United Kingdom,Clin Chim Acta 309 (2001),179–181.

[15] J.J. Going and B.A. Gusterson, Molecular pathology and fu-ture developments,Eur J Cancer 35 (1999), 1895–1904.

[16] Guidance for Industry, Bioanalytical Method Validation, USDept of Health and Human Services, Food and Drug Adminis-tration, Center for Drug Evaluation and Research, May 2001,pp. 22.

[17] W.B. Hamlin, Requirements for accreditation by the Collegeof American Pathologists Laboratory Accreditation Program,Arch Pathol Lab Med 123 (1999), 465–467.

[18] C. Hartmann, D.L. Massart and R.D. McDowall, An analysisof the Washington Conference Report on bioanalytical methodvalidation,J Pharm Biomed Anal 11 (1994), 1337–1343.

[19] C. Hartmann, J. Smeyers-Verbeke, D.L. Massart and R.D. Mc-Dowall, Validation of bioanalytical chromatographic methods,J Pharm Biomed Anal 17 (1998), 193–218.

[20] P. Hess and D. Cooper, Impact of pharmacogenomics on theclinical laboratory,Mol Diagn 4 (1999), 289–298.

[21] E. Hofmann, What can medicine learn from the human DNAsequence?Biochem 66 (2001), 1144–1152.

[22] W. Huisman, European Communities Confederation of Clini-cal Chemistry Working Group on Accreditation: past, presentand future,Clin Chim Acta 309 (2001), 111–114.

[23] International Conference on Harmonisation of Technical Re-quirements for Registration of Pharmaceuticals for HumanUse, ICH Harmonized Tripartite Guideline, Text on validationof analytical procedures, 1994.

[24] International Conference on Harmonisation of Technical Re-quirements for Registration of Pharmaceuticals for HumanUse, ICH Harmonized Tripartite Guideline, Validation of an-alytical procedures: methodology, 1996.

[25] K.M. Irjala and P.E. Gronroos, Preanalytical and analyticalfactors affecting laboratory results,Ann Med 30 (1998), 267–272.

[26] D. Kenny, ISO and CEN documents on quality in medicallaboratories,Clin Chim Acta 309 (2001), 121–125.

[27] J. Kristiansen, Description of a generally applicable modelfor the evaluation of uncertainty of measurement in clinicalchemistry,Clin Chem Lab Med 10 (2001), 920–931.

[28] U. Landegren, Consulting the source code: prospects for gene-based medical diagnostics,J Intern Med 248 (2000), 271–276.

[29] Z. Laszlo, A. Rossler and H.G. Hinghofer-Szalkay, Cardio-vascular and humoral readjustment after different levels ofhead-up tilt in humans,Aviat Space Environ Med 72 (2001),193–201.

[30] J.C. Libeer, Role of external quality assurance schemes inassessing and improving quality in medical laboratories,ClinChim Acta 309 (2001), 173–177.

[31] C. Kluft, A.f. Jie, D.C. Rijkin and J.H. Verheijen, Daytimefluctuations in blood tissue-type plasminogen activator (t-PA)and its fast acting inhibitor (PAI-1),Thromb Haemost 59(1988), 329–332.

[32] D.H. Manicourt, P. Poilvache, A. Nzeusseu, A. van Egeren, J.P.Devogelaer, M.E. Lenz and E.J.-M.A. Thonar, Serum levelsof hyaluronan, antigenic keratan sulfate, matrix metallopro-teinase 3, and tissue inhibitor of metalloproteinase 1, changepredictably in rheumatoid arthritis patients who have begunactivity after a night of bed rest,Arthritis Rheum 42 (1999),1861–1869.

[33] T.A. Manolio, G.L. Burke, P.J. Savage, D.R. Jacobs, Jr, S.Sidney, L.E. Wagenknecht, R.M. Allman and R.P. Tracy, Sexand race-related differences in liver-associated serum chem-istry tests in young adults in the CARDIA study,Clin Chem38 (1992), 1853–1859.

[34] S. Narayanan, The preanalytic phase. An important componentof laboratory medicine,Am J Clin Pathol 113 (2000), 429–452, Comment in:Am J Clin Pathol. 114 (2000), 820–821.

[35] M.A. Owen, H.G. Vall, A.A. Hurley and S.B. Wormsley, Val-idation and quality control of immunophenotyping in clinicalflow cytometry,J Immunol Meth 243 (2000), 33–50.


[36] F. Podo, O. Henriksen, W.M. Bovee, M.O. Leach, D. Leibfritzand J.D. de Certaines, Absolute metabolite quantification byin vivo NMR spectroscopy: I. Introduction, objectives andactivities of a concerted action in biomedical research,MagnReson Imaging 16 (1998), 1085–1092.

[37] C. Ricos, V. Alvarez, F. Cava, J.V. Garcia-Lario, A. Her-nandez, C.V. Jimenez, J. Minchinela, C. Perich and M. Si-mon, Current databases on biological variation: pros, consand progress,Scand J Clin Lab Invest 59 (1999), 491–500.

[38] K.W. Ryder and M.R. Glick, Erroneous laboratory results fromhemolyzed, icteric, and lipemic specimens,Clin Chem 39(1993), 175–176.

[39] C. Sander, Genomic medicine and the future of health care,Science 287 (2000), 1977–1988.

[40] L. Sciacovelli, S. Secchiero, L. Zardo and M. Plebani, Ex-ternal Quality Assessment Schemes: need for recognised re-quirements,Clin Chim Acta 309 (2001), 183–199.

[41] V.P. Shah, K.K. Midha, S. Dighe, I.J. McGilveray, J.P. Skelly,A. Yacobi, T. Layloff, C.T. Viswanathan, C.E. Cook, R.D.McDowall, K.A. Pittman and S. Spector, Analytical methodsvalidation: Bioavailability, bioequivalence, and pharmacoki-netic studies. Conference Report,Europ J. Drug Metabolismand Pharmacokinetics 16 (1991), 249–255.

[42] V.P. Shah, K.K. Midha, J.W.A. Findlay, H.M. Hill, J.D. Hulse,I.J. McGilveray, G. McKay, K.J. Miller, R.N. Patnaik, M.L.

Powell, A. Tonelli, C.T. Viswanathan and A. Yacobi, Bioana-lytical method validation, a revisit with a decade of progress,Pharm Res 17 (2000), 1551–1557.

[43] M.M. Shi, Enabling large-scale pharmacogenetic studies byhigh-throughput mutation detection and genotyping technolo-gies,Clin Chem 47 (2001), 164–172.

[44] S. Slagter and J.G. Loeber, Accreditation of medical laborato-ries in The Netherlands,Clin Chim Acta 309 (2001), 155–161.

[45] G. Sterner, M. Wroblewski and U. Rosen, Postprandial in-crease in serum creatinine in renal transplant recipients,Transpl Int 5 (1992), 115–117.

[46] L.M. Szymanski and R.R. Pate, Effects of exercise intensity,duration, and time of day on fibrinolytic activity in physicallyactive men,Med Sci Sports Exerc 26 (1994), 1102–1108.

[47] N.W. Tietz, D.F. Shuey and D.R. Wekstein, Laboratory valuesin fit aging individuals – sexagenarians through centenarians,Clin Chem 38 (1992), 1167–1185.

[48] N. Wedemeyer and T. Potter, Flow cytometry: an “old” tool fornovel applications in medical genetics,Clin Genet 60 (2001),1–8.

[49] J.O. Westgard,Basic Method Validation, AACC Press, Wash-ington, DC, 1999, pp. 256.

[50] D.S. Young,Effects of Drugs on Clinical Laboratory Tests,(Fourth ed.), AACC Press, Washington, DC, 1995, pp. 1462.

Submit your manuscripts athttp://www.hindawi.com

Stem CellsInternational

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Immunology ResearchHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinson’s Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttp://www.hindawi.com

delivery of high-quality biomarker assaysdownloads.hindawi.com/journals/dm/2002/212987.pdf ·...

Documents